ZACH Insights · Genetics 101
Genetics 101.
DNA, traits
and training.
How cells use genetic information, why people differ and how to interpret genetic findings.
Begin with the cell ↓01 · Build the picture
Begin with a cell.
Then change the scale.
A muscle can contract, a neuron can send a signal and a liver cell can process nutrients. These tasks require different cellular machinery. Genetics helps explain how cells build and regulate that machinery. It does not make the rest of biology disappear.
A cell is a small living unit. Many human cells contain a nucleus, the compartment holding most of their DNA. There are exceptions: mature red blood cells lack a nucleus, and mitochondria contain their own small genome. The diagram below is an introductory view of nuclear DNA.
DNA, or deoxyribonucleic acid, is a molecule whose sequence carries biological information. Chromosomes organise long DNA molecules together with associated proteins. Most human body cells have 23 pairs of nuclear chromosomes; reproductive cells normally carry one set. A chromosome is not a separate substance wrapped around DNA—it includes the DNA itself. [1]
The living setting in which genetic information is used.
A long DNA molecule packaged with proteins.
Paired strands whose base sequence carries information.
A sugar, a phosphate group and a base.
The letters are chemical parts.
DNA is built from nucleotides. Each contains a sugar called deoxyribose, a phosphate group and one base. Sugar and phosphate form the strand's backbone. The bases project from it, pairing across the two strands: adenine with thymine, cytosine with guanine. The familiar letters A, T, C and G name those bases. [2]
Think of a letter in a book. Its shape and position matter, but it still belongs to a physical page. Similarly, a DNA base carries meaning through its place in a sequence while remaining part of a chemical structure. The analogy explains sequence; DNA is not literally a written instruction manual.
02 · Information becomes activity
A gene is used.
A function emerges.
A gene is a DNA sequence contributing to a functional product. Some genes encode proteins. Others produce RNA that functions without becoming a protein. The shortcut “one gene makes one protein” therefore misses both non-coding RNA and the ways one gene can contribute to different products. [3]
Gene expression means using a gene's information to produce RNA, and for protein-coding genes, protein. Expression is regulated: a cell does not use every gene equally at every moment. The presence of a gene, the amount of its RNA and the amount or activity of a resulting protein are distinct observations.
A cell builds an RNA copy using a DNA template. The sequence is transcribed into another nucleic acid.
For many human protein-coding genes, RNA is processed before use. Different combinations of included segments can produce different messages.
A ribosome reads the message in groups of three bases and assembles a chain of amino acids.
An amino acid is a building block of proteins. The chain's sequence influences how it folds and functions. Proteins can form structures, catalyse chemical reactions, transport substances or help transmit signals. A muscle's behaviour depends on many such components working together—not one “strength gene”.
Messenger RNA(mRNA) carries information for protein production. Transfer RNA(tRNA) helps bring amino acids to the ribosome. Ribosomal RNA(rRNA) is part of the machinery doing the work. RNA is therefore more than a temporary copy of DNA. [4]
How are DNA and RNA different?
DNA contains deoxyribose and the bases A, T, C and G. RNA contains ribose and the bases A, U, C and G. RNA is often single-stranded but can fold into complex structures. “DNA stays in the nucleus; RNA leaves it” is an introductory picture of nuclear gene expression, not a universal rule for all cellular biology.
Why can different cells behave differently with similar DNA?
Cells differ in which genes they express, when they express them and how their products are processed and regulated. Their surroundings and signals also differ. Possessing similar nuclear DNA does not mean having the same active machinery.
The pathway above shows protein-coding expression. Functional RNA is another endpoint. Arrows indicate an information pathway; they do not promise a simple one-to-one effect on a person's traits.
03 · Similar structures, different sequences
A variant is a difference.
Its meaning needs context.
A genetic variant is a difference in DNA sequence. Differences range from a single base to larger changes involving segments of DNA. A variant can alter a product, affect regulation or have no established functional effect. “Different” is not synonymous with “damaging”. [5]
An allele is a version at a genetic location. Your genotype describes the versions you carry. A phenotype is an observable or measurable characteristic, such as a laboratory measurement or height. For complex characteristics, the path from genotype to phenotype passes through many biological and environmental influences.
Keep those levels separate. Measuring a variant is not the same as measuring a protein, a trait or a future outcome. A DNA result needs evidence connecting the particular variant to the particular question.
A deliberately simple sequence example
A C G A T C
A C G G T C
Many small associations are still not a personal explanation.
A genome-wide association study, usually shortened to GWAS, compares genetic markers and a measured trait across many people. It asks which variants are statistically associated with differences in that trait. An associated marker may point towards a biologically relevant region, but association alone does not identify the complete causal pathway. [6]
A polygenic index combines information across many variants using weights estimated in a research sample. Its usefulness depends on the outcome, measurement, study design and population where it is applied. A population association can be informative while prediction for one person remains limited. “Statistically detectable” and “accurately predictive” are different achievements.
04 · A statement about differences between people
Heritability belongs
to a population.
Heritability is a statistical description of how genetic differences relate to variation in a trait in a particular population under particular conditions. It is not the percentage of one person's characteristic “made by genes”. Nor does it tell us how easily that characteristic can change. [7]
Imagine two groups growing plants. In one group, light and water are very similar. In the other, those conditions vary substantially. Genetic differences could be similar in both groups, yet the share of observed variation associated with them could differ. The comparison concerns differences across plants. It does not divide an individual plant into a genetic part and an environmental part.
The figure translates that idea into a deliberately simplified additive model. Real human traits can involve interactions, correlations, measurement error and more complex processes. The figure is a teaching analogy, not an estimate from data.
Conceptual model · Not observed data
When environments vary less, genetic differences can account for a larger share of variation in this simple model. The genetic row has not changed.
Twin and family methods
These methods compare patterns of resemblance among relatives. Their interpretation depends on assumptions about genetic relatedness, shared environments and the process producing resemblance.
Methods using measured variants
SNP heritability estimates variation associated with the genetic variants indexed by a particular molecular method. It does not necessarily capture the same influences as a twin study. Different methods can therefore give different estimates.
A SNP is a single-nucleotide polymorphism, a form of single-base variation. SNP heritability, overall heritability and the predictive accuracy of a polygenic index are related concepts with different meanings.
A concrete research example · Personality
Large studies sharpen
the question.
1.14 million Participants per Big Five trait in the main GWAS analyses.
Schwaba and colleagues studied the Big Five personality traits: extraversion, agreeableness, conscientiousness, neuroticism and openness. They combined cohorts and used within-family analyses to examine the robustness of genetic associations. [8]
For the primary population-level analyses with European-like genomes, estimated SNP heritability ranged from 4.8% to 9.3% across traits. Estimates changed with measurement reliability and analytic approach. These figures describe variation in measured traits, not how much of a person's personality is genetic.
The paper discusses larger estimates from other methods. The difference does not reduce to one estimate being “the genetic percentage”. Methods capture different components and rely on different assumptions. Results also need attention to ancestry, measurement and the limits of generalisation.
The practical reading lesson is to ask which trait, which sample, which method and which estimate before repeating a number.
05 · Bring the concepts back to a training week
Biological variation
leaves room for observation.
Two people may show different changes after what appears to be the same training programme. Genetics is one possible contributor. That observation alone cannot tell us how much the difference is genetic—or even whether the programme delivered the same effective training dose.
Consider a simple example. Two runners are given three weekly sessions. One completes every session; the other misses several because of work. They begin with different fitness, interpret “easy” differently and are tested on days with different fatigue. Their measured changes now include differences in exposure, starting point and measurement. Calling one a “genetic non-responder” would jump beyond the evidence.
That is why practical adjustment begins with things we can observe: the work actually completed, its progression, recovery, relevant symptoms and repeated outcomes measured in a consistent way. Individual variation justifies paying attention. It does not automatically justify a DNA-based prescription.
The FIMS consensus update published in 2020 described major limits to the use of genetic information for predicting athletic performance and talent identification. That is a dated consensus, not a claim that research can never progress. Any proposed test still needs evidence for its intended use—including evidence that acting on it improves the outcome that matters. [9]
Epigenetics is regulation, not a slogan.
Epigenetic mechanisms influence how DNA is used without changing its underlying sequence. Chemical modifications to DNA and associated proteins can be part of that regulation. They vary across tissues and biological conditions. [10]
A change in an epigenetic mark is not, on its own, evidence of an improved life outcome. It also does not mean a thought or one exercise has “rewritten your DNA”. Follow the chain: what changed, where it was measured, whether it altered function and whether that function matters to the question.
A useful sequence of decisions.
- Define the ability or habit you want to develop.
- Choose a feasible training structure and record what you complete.
- Compare repeated, relevant observations rather than one unusually good or bad day.
- Adjust the programme in light of those observations and appropriate professional guidance.
This is a learning framework, not an individual treatment plan or a genetic service.
Connect this to strength and endurance →06 · Read a genetic report
From a measured variant
to a useful decision.
A report can contain precise laboratory measurements alongside much less certain interpretations. Read those layers separately. The following sequence helps you locate what has been measured, what has been inferred and what evidence a proposed action still needs.
What was detected?
Which variants or regions were tested? How accurately does the method identify them? A result from selected markers is not a complete reading of every possible genetic influence.
What does research link it to?
Which trait was measured, in which population and with what effect size? An association with a trait is a research finding; it does not yet tell you how accurately an individual can be predicted.
How informative is it for a person?
Was the prediction tested in an independent sample relevant to the intended user? A ranking or percentile is not automatically a probability, a diagnosis or a measured physical ability.
Does acting on it help?
A useful recommendation needs more than a detectable association. Ask whether choosing care or training based on the result improves a meaningful outcome compared with an appropriate alternative.
The first distinction concerns analytical validity: whether the test measures what it claims to measure. The next concerns clinical validity: whether the measurement relates to the relevant health state. Those are distinct from the practical value of using a result to guide a decision. [11]
A genetic score needs a reference.
A polygenic score places a collection of genetic associations in context. To interpret it, ask who supplied the reference data and what the reported scale means. A position relative to others is not the same as an absolute risk over a specified period. Age, family history and other relevant factors can change the interpretation. [12]
For training, a DNA-based label and a measured result answer different questions. A laboratory exercise test measures oxygen uptake under specified conditions. A genetic report may estimate a predisposition using associations. One cannot simply replace the other.
Questions worth asking.
Does a genetic influence make a trait fixed?
No. Heritability concerns differences in a population under particular conditions. It does not measure how much a person can change. Genetic information and the effects of training, development and surroundings belong in the same biological picture.
Does a lower genetic risk mean no risk?
No. Consumer reports may examine only some relevant variants, and other factors contribute to health. A lower-risk or negative result does not rule out an outcome. A higher-risk result does not mean it will certainly happen. Health-related findings deserve interpretation with a qualified healthcare or genetics professional. [13]
Does a sports report reveal my ideal programme?
That conclusion requires evidence for the specific test and intended use. Ask whether its predictions are accurate and whether following its recommendation improves an outcome. Continue to observe the training actually completed, recovery and measured performance rather than treating a predisposition label as a verdict.
What should a trustworthy explanation show?
The measured variants, the research sources, the studied population, the meaning of the scale and the limitations. It should separate measurement from interpretation and make clear which proposed actions have supporting evidence.
Keep the vocabulary within reach
A glossary you can return to.
- Allele
- A version at a genetic location. The term describes which version is present, not its effect by itself.
- Amino acid
- A building block of proteins. An amino-acid sequence becomes part of a molecule with a particular structure and function.
- Chromosome
- A DNA molecule organised with associated proteins. Most human body cells contain 23 pairs of nuclear chromosomes.
- DNA
- Deoxyribonucleic acid, a molecule carrying genetic information through its base sequence.
- Epigenetics
- Mechanisms affecting the use of DNA without changing its underlying sequence.
- Gene
- A DNA sequence contributing to a functional product, including a protein or functional RNA.
- Gene expression
- The production of RNA, and for protein-coding genes, protein, using genetic information.
- Genetic variant
- A difference in DNA sequence. Its significance depends on the particular difference and the evidence.
- Genotype
- The genetic versions an individual carries at one or more locations.
- GWAS
- Genome-wide association study: research comparing genetic markers with measured traits across many people.
- Heritability
- A population statistic concerning the contribution of genetic differences to variation in a trait under specified conditions.
- Nucleotide
- A nucleic-acid building block containing a sugar, a phosphate group and a base.
- Phenotype
- An observable or measurable characteristic arising through biological development and context.
- Polygenic index
- A weighted combination of information from multiple genetic variants. Predictive usefulness needs validation for the intended population and outcome.
- RNA
- Ribonucleic acid. Its forms include messenger, transfer and ribosomal RNA, with distinct cellular roles.
- SNP heritability
- Variation associated with the genetic variants indexed by a molecular method. It is not identical to an overall heritability estimate or personal predictive accuracy.
Sources & scope
Follow the evidence.
The diagrams and teaching examples are original conceptual illustrations. They are not patient data. This article introduces concepts and a selected research example; it is not a systematic review or individual genetic counselling.
- NHGRI. DNA fact sheet. Nuclear DNA and chromosomes.
- NHGRI. Nucleotide. Components of a nucleotide.
- NHGRI. Gene and gene expression.
- NHGRI. RNA fact sheet. Expression, RNA roles and alternative splicing.
- MedlinePlus Genetics. Gene variants; NHGRI. Human genomic variation.
- NHGRI. Genome-wide association studies.
- MedlinePlus Genetics. What is heritability? Population scope and interpretation.
- Schwaba T, Clapp Sullivan ML, Akingbuwa WA and colleagues (2026). Robust inference and correlates from genetic associations with personality. Nature. DOI 10.1038/s41586-026-10992-9. Selected findings checked against the primary article.
- Tanisawa K and colleagues (2020). Sport and exercise genomics: the FIMS 2019 consensus statement update. British Journal of Sports Medicine. Used for its discussion of prediction limits, not as an individual prescription.
- NHGRI. Epigenomics fact sheet. Regulation and underlying DNA sequence.
- FDA. Direct-to-consumer tests. Test validity and interpretation; US regulatory information is not presented here as Danish regulation.
- NHGRI. Polygenic risk scores. Relative prediction, reference populations and context.
- MedlinePlus Genetics. What do consumer genetic test results mean?. Coverage and interpretation limits.
Current ZACH services concern movement, training and education. Genetic testing, diagnosis and individual genetic interpretation are not offered through this article.